A. Field of the Invention
The present invention relates generally to nucleic acid sequences useful as probes for the diagnosis of cancer and methods relating thereto. More particularly, the present invention concerns probes and methods useful in diagnosing, identifying and monitoring the progression of prostate cancer, benign prostatic hyperplasia, bladder cancer or breast cancer through measurements of gene products.
B. Description of the Related Art
Genetic detection of human disease states is a rapidly developing field (Taparowsky et al., 1982; Slamon et al., 1989; Sidransky et al., 1992; Miki et al., 1994; Dong et al., 1995; Morahan et al., 1996; Lifton, 1996; Barinaga, 1996). However, some problems exist with this approach. A number of known genetic lesions merely predispose to development of specific disease states. Individuals carrying the genetic lesion may not develop the disease state, while other individuals may develop the disease state without possessing a particular genetic lesion. In human cancers, genetic defects may potentially occur in a large number of known tumor suppresser genes and proto-oncogenes.
The genetic detection of cancer has a long history. One of the earliest genetic lesions shown to predispose to cancer was transforming point mutations in the ras oncogenes (Taparowsky et al., 1982). Transforming ras point mutations may be detected in the stool of individuals with benign and malignant colorectal tumors (Sidransky et al, 1992). However, only 50% of such tumors contained a ras mutation (Sidransky et al., 1992). Similar results have been obtained with amplification of HER-2/neu in breast and ovarian cancer (Slamon et al., 1989), deletion and mutation of p53 in bladder cancer (Sidransky et al., 1991), deletion of DCC in colorectal cancer (Fearon et al., 1990) and mutation of BRCA1 in breast and ovarian cancer (Miki et al., 1994).
None of these genetic lesions are capable of predicting a majority of individuals with cancer and most require direct sampling of a suspected tumor, making screening difficult.
Further, none of the markers described above are capable of distinguishing between metastatic and non-metastatic forms of cancer. In effective management of cancer patients, identification of those individuals whose tumors have already metastasized or are likely to metastasize is critical. Because metastatic cancer kills 560,000 people in the US each year (ACS home page), identification of markers for metastatic cancer, such as metastatic prostate and breast cancer, would be an important advance.
A particular problem in cancer detection and diagnosis occurs with prostate cancer. Carcinoma of the prostate (PCA) is the most frequently diagnosed cancer among men in the United States (Veltri et al., 1996). Prostate cancer was diagnosed in approximately 210,000 men in 1997 and about 39,000 men succumbed to the malignancy (Parker et al., 1996; Wingo et al., 1997). The American Cancer Society expects that more than 340,000 new cases of prostate cancer will be diagnosed in 1998 (Orozco et al., 1998). Although relatively few prostate tumors progress to clinical significance during the lifetime of the patient, those which are progressive in nature are likely to have metastasized by the time of detection. Survival rates for individuals with metastatic prostate cancer are quite low. Between these extremes are patients with prostate tumors that will metastasize but have not yet done so, for whom surgical prostate removal is curative. Determination of which group a patient falls within is critical in determining optimal treatment and patient survival.
The FDA approval of the serum prostate specific antigen (PSA) test in 1984 has subsequently changed the way prostate disease was managed (Allhoff et al., 1989; Cooner et al., 1990; Jacobson et al., 1995; Orozco et al., 1998). PSA is widely used as a serum biomarker to detect and monitor therapeutic response in prostate cancer patients (Badalament et al., 1996; O'Dowd et al, 1997). Several modifications in PSA assays (Partin and Oesterling, 1994; Babian et al., 1996; Zlotta et al., 1997) have resulted in earlier diagnoses and improved treatment.
While an effective indicator of prostate cancer when serum levels are relatively high, PSA serum levels are more ambiguous indicators of prostate cancer when only modestly elevated, for example when levels are between 2-10 ng/ml. At these modest elevations, serum PSA may have originated from non-cancerous disease states such as BPH (benign prostatic hyperplasia), prostatitis or physical trauma (McCormack et al., 1995). Although application of the lower 2.0 ng/ml cancer detection cutoff concentration of serum PSA has increased the diagnosis of prostate cancer, especially in younger men with non-palpable early stage tumors (Stage Tlc) (Soh et al., 1997; Carter et al., 1997; Harris et al., 1997; Orozco et al., 1998), the specificity of the PSA assay for prostate cancer detection at low serum PSA levels remains a problem.
In current clinical practice, the serum PSA assay and digital rectal exam (DRE) is used to indicate which patients should have a prostate biopsy (Lithrup et al., 1994; Orozco et al., 1998). Histological examination of the biopsied tissue is used to make the diagnosis of prostate cancer. Based upon the American Cancer Society estimate of 340,000 cases of diagnosed prostate cancer in 1998 and a known cancer detection rate of about 35% (Parker et al., 1996), it is estimated that in 1998 over one million prostate biopsies will be performed in the United States (Orozco et al., 1998). Clearly, there would be much benefit derived from a serological test that was sensitive enough to detect small and early stage prostate tumors that also had sufficient specificity to exclude a greater portion of patients with noncancerous or clinically insignificant conditions.
Several investigators have sought to improve upon the specificity of serologic detection of prostate cancer by examining a variety of other biomarkers besides serum PSA concentration (Ralph and Veltri, 1997). One of the most heavily investigated of these other biomarkers is the ratio of free versus total PSA (f/t PSA) in a patient's blood. Most PSA in serum is in a molecular form that is bound to other proteins such as .alpha.1-antichymotrypsin (ACT) or .alpha.2-macroglobulin (Christensson et al., 1993; Stenman et al., 1991; Lilja et al., 1991). Free PSA is not bound to other proteins. The ratio of free to total PSA (f/tPSA) is usually significantly higher in patients with BPH compared to those with organ confined prostate cancer (Marley et al., 1996; Oesterling et al., 1995; Pettersson et al., 1995). When an appropriate cutoff is determined for the f/tPSA assay, the f/tPSA assay can help distinguish patients with BPH from those with prostate cancer in cases in which serum PSA levels are only modestly elevated (Marley et al., 1996; Partin and Oesterling, 1996). Unfortunately, while f/tPSA may improve on the detection of prostate cancer, information in the f/tPSA ratio is insufficient to improve the sensitivity and specificity of serologic detection of prostate cancer to desirable levels.
Genetic changes reported to be associated with prostate cancer include: allelic loss (Bova, et al., 1993; Macoska et al., 1994; Carter et al., 1990); DNA hypermethylation (Isaacs et al., 1994); point mutations or deletions of the retinoblastoma (Rb) and p53 genes (Bookstein et al., 1990a; Bookstein et al., 1990b; Isaacs et al., 1991); and aneuploidy and aneusomy of chromosomes detected by fluorescence in situ hybridization (FISH) (Macoska et al., 1994; Visakorpi et al., 1994; Takahashi et al., 1994; Alcaraz et al., 1994).
A recent development in this field was the identification of a prostate metastasis suppresser gene, KAI1 (Dong et al., 1995). Insertion of wild-type KAI1 gene into a rat prostate cancer line caused a significant decrease in metastatic tumor formation (Dong et al., 1995). However, detection of KAI1 mutations is dependent upon direct sampling of mutant prostate cells. Thus, either a primary prostate tumor must be sampled or else sufficient transformed cells must be present in blood, lymph nodes or other tissues to detect the missing or abnormal gene. Further, the presence of a deleted gene may frequently be masked by large numbers of untransformed cells that may be present in a given tissue sample.
The most commonly utilized current tests for prostate cancer are digital rectal examination (DRE) and analysis of serum prostate specific antigen (PSA). Although PSA has been widely used as a clinical marker of prostate cancer since 1988 (Partin & Oesterling, 1994), screening programs utilizing PSA alone or in combination with digital rectal examination have not been successful in improving the survival rate for men with prostate cancer (Partin & Oesterling, 1994). While PSA is specific to prostate tissue, it is produced by normal and benign as well as malignant prostatic epithelium, resulting in a high false-positive rate for prostate cancer detection (Partin & Oesterling, 1994).
Other markers that have been used for prostate cancer detection include prostatic acid phosphatase (PAP) and prostate secreted protein (PSP). PAP is secreted by prostate cells under hormonal control (Brawn et al., 1996). It has less specificity and sensitivity than does PSA. As a result, it is used much less now, although PAP may still have some applications for monitoring metastatic patients that have failed primary treatments. In general, PSP is a more sensitive biomarker than PAP, but is not as sensitive as PSA (Huang et al., 1993). Like PSA, PSP levels are frequently elevated in patients with BPH as well as those with prostate cancer.
Another serum marker associated with prostate disease is prostate specific membrane antigen (PSMA) (Horoszewicz et al., 1987; Carter et al., 1996; Murphy et al., 1996). PSMA is a Type II cell membrane protein and has been identified as Folic Acid Hydrolase (FAH) (Carter et al., 1996). Antibodies against PSMA react with both normal prostate tissue and prostate cancer tissue (Horoszewicz et al., 1987). Murphy et al. (1995) used ELISA to detect serum PSMA in advanced prostate cancer. As a serum test, PSMA levels are a relatively poor indicator of prostate cancer. However, PSMA may have utility in certain circumstances. PSMA is expressed in metastatic prostate tumor capillary beds (Silver et al., 1997) and is reported to be more abundant in the blood of metastatic cancer patients (Murphy et al., 1996). PSMA messenger RNA (mRNA) is down-regulated 8-10 fold in the LNCaP prostate cancer cell line after exposure to 5-.alpha.-dihydroxytestosterone (DHT) (Israeli et al., 1994).
A relatively new potential biomarker for prostate cancer is human kallekrein 2 (HK2) (Piironen et al., 1996). HK2 is a member of the kallekrein family that is secreted by the prostate gland. In theory, serum concentrations of HK2 may be of utility in prostate cancer detection or diagnosis, but the usefulness of this marker is still being evaluated.
As prostate cancer is one of the most prevalent forms of cancer in men, breast cancer is one of the most prevalent forms of cancer in women. Breast cancer is the leading cause of death for women between 30-50 years of age in the United States. Pathological breast cancer staging (tumor size, nodal status) is still the most reliable method for predicting outcome. In contrast to other forms of cancer, only a few tumor markers have been identified for breast cancer (e.g., estrogen receptor, progesterone receptor, S-phase, P53, Erb-2, cathepsin D) (see, e.g. Slamon el al, 1987).
Mutational analysis of important tumor suppressor genes such as p53 (Elledge, 1994) and BRCA1 (Miki et al., 1994) has recently been introduced as a diagnostic and prognostic test for breast cancer. However, many of those markers are not reliable enough to be used for routine purposes in the clinic. Two tumor suppressor genes that are mutated in a number of other cancers (Rb and p53) show a frequency of mutation of only about 30% in breast cancer (Cox et al., 1994). Mutations in the recently identified breast cancer susceptibility genes BRCA1 (chromosome 17q21) and BRCA2 (chromosome 13q13) are associated with familial breast cancer, accounting for about 5% of total breast cancer cases, but have not been found in sporadic breast cancer (Stratton and Wooster, 1996). There has yet to be found a single genetic change that accounts for the majority of sporadic breast cancers. Therefore, there is an urgent need for better prognostic markers in breast cancer diagnosis.
It is known that the processes of transformation and tumor progression are associated with changes in the levels of messenger RNA species (Slamon et al., 1984; Sager et al., 1993; Mok et al., 1994; Watson et al., 1994). Recently, a variation on PCR analysis known as RNA fingerprinting has been used to identify messages differentially expressed in ovarian or breast carcinomas (Liang et al., 1992; Sager et al., 1993; Mok et al., 1994; Watson et al., 1994). By using arbitrary primers to generate "fingerprints" from total cell RNA, followed by separation of the amplified fragments by high resolution gel electrophoresis, it is possible to identify RNA species that are either up-regulated or down-regulated in cancer cells. Results of these studies indicated the presence of several markers of potential utility for diagnosis of breast or ovarian cancer, including a6-integrin (Sager et al., 1993), DEST001 and DEST002 (Watson et al., 1994), and LF4.0 (Mok et al., 1994).
There remain, however, deficiencies in the prior art with respect to the identification of the genes linked with the progression of prostate, bladder or breast cancer and the development of diagnostic methods to monitor disease progression. Likewise, the identification of genes which are differentially expressed in prostate, bladder, breast and other forms of cancer would be of considerable importance in the development of a rapid, inexpensive method to diagnose cancer.